Cocaine dependence is a significant ongoing public health concern, in part because there is no effective pharmacological treatment, and the likelihood of relapse after recovery remains high. Cocaine is known to cause long-term, persistent changes in the prefrontal cortex, a brain region responsible for decision-making strategies and reward valuation. These changes occur even at the level of dendritic spines, the primary communication points between neurons. Cocaine-induced alteration of prefrontal cortical dendritic spines may thus be a contributing factor to cocaine addiction and relapse. Both humans and rodents can learn to associate specific actions with their outcomes. These actions are considered 'goal-directed,' meaning their performance is sensitive to changes in the relationship between the action and its outcome. With repetition or exposure to drugs of abuse such as cocaine, goal-directed responding can shift from being outcome-sensitive to being automatic, or 'habitual', which is defined by insensitivity to changes in the action-outcome relationship. Restoring goal-directed decision-making after habit formation in experimental models has proven particularly challenging, but it is a critical research imperative because although the development of habits can serve adaptive purposes, habitual behaviors that occur at the expense of goal-directed decision-making are thought to play a role in addiction etiology. This proposal aims to identify the cellular mechanisms of goal-directed behavior and to restore goal-directed decision-making in mice that have developed habits due to cocaine exposure. I will focus on ROCK, a key regulator of the actin cytoskeleton that controls contractile-dependent remodeling of dendritic spines. I have shown that the ROCK inhibitor HA-1077 restores goal-directed decision-making in extensively trained mice that otherwise engage in stimulus-response habits. By contrast, cocaine exposure induces habits. These findings raise the provocative possibility that targeting cytoskeletal regulators such as ROCK may be useful in the treatment of cocaine dependence and prevention of relapse. I propose to: 1) Reverse cocaine-induced habits, as well as block habitual responding for cocaine, using HA-1077. 2) Use in vivo multiphoton microscopy to quantify, in real time, the morphological response of prefrontal cortical dendritic spines to cocaine exposure. I will next isolate the effects of HA-1077 intervention with the expectation that HA-1077 will reverse the structural consequences of cocaine exposure. Whether cocaine-induced alterations of prefrontal cortical dendritic spines are a contributing factor to addiction etiology represents a lively debate in the field, especially since few research groups are equipped with the tools to both model cocaine-induced maladaptive decision-making and to isolate cytostructural modifications in tandem. I will harness these tools to comprehensively address this gap in current knowledge.